JevMade hello@JevMade.com
← Back to experiments

Benchmarks & research

jev-measured

Measures what the live Jev API actually returns across eight realistic use cases, reporting cost, latency and response shape rather than rate-card arithmetic.

Source screenshot of jev-measured
SOURCE SCREENSHOT · source ↗ · captured 2026-09-24Full screenshot ↗

What it does

Three results a builder should know: a real decision costs $0.0000153–$0.0000226, Score returns a probability-weighted rubric index rather than a 0–1 float, and Noul carries no confidence field.

Maker-reported (not independently measured by JevMade): rag-reranking 364 input tokens, cost $0.0000153; support-ticket-triage 539 input tokens, $0.0000226 · Pricing is $0.042 per 1M input tokens with no output tokens billed · Raw wall-clock medians 458-563 ms; measured TCP handshake floor median 198.8 ms (7 samples, 191.2-266.4 ms)

Primitives
choice, score, noul
Platform
Python
Added
Project created

Source checked 2026-09-24 — opened the GitHub repository directly.